cloud-fan commented on PR #59271:
URL: https://github.com/apache/spark/pull/59271#issuecomment-6171259351
Thanks for the rework. Merging the nested nullability on the Expand output
with StructType.unionLikeMerge, as Union.mergeChildOutputs does, removes the
nullability-only casts and keeps the first value's nested field names and
metadata. The new DatasetUnpivotSuite test checks both value orders and that
the analyzed plan has no Cast. UnpivotResolver now checks value sizes first and
coerces only when valuesTypeCoercioned is false, matching UnpivotCoercion, so
the legacy flag behaves the same in both analyzers. As you noted,
UnpivotResolver is not dispatched by the OSS single-pass Resolver, so that path
is checked by reading the code rather than by an end-to-end test.
<!-- SPARK_DEV_REVIEW_REPLY
{"feedback_id":"conversation:6148899414","thread_id":"conversation:6148899414","verdict_sha256":"e3e39896ac52c941c2b907dc5df4a84120fadb433dd185df928cabb8fa9d4d43"}
-->
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
